2000-03-31 22:58:54 +08:00
|
|
|
#
|
|
|
|
# Secret Labs' Regular Expression Engine
|
|
|
|
#
|
|
|
|
# convert template to internal format
|
|
|
|
#
|
|
|
|
# Copyright (c) 1997-2000 by Secret Labs AB. All rights reserved.
|
|
|
|
#
|
|
|
|
# Portions of this engine have been developed in cooperation with
|
2000-07-02 01:50:59 +08:00
|
|
|
# CNRI. Hewlett-Packard provided funding for 2.0 integration and
|
2000-03-31 22:58:54 +08:00
|
|
|
# other compatibility work.
|
|
|
|
#
|
|
|
|
|
|
|
|
import _sre
|
|
|
|
|
|
|
|
from sre_constants import *
|
|
|
|
|
2000-07-02 20:00:07 +08:00
|
|
|
MAXCODE = 65535
|
|
|
|
|
2000-07-06 05:14:16 +08:00
|
|
|
def _charset(charset, fixup=None):
|
2000-07-02 20:00:07 +08:00
|
|
|
# internal: optimize character set
|
2000-07-06 05:14:16 +08:00
|
|
|
if not fixup:
|
|
|
|
fixup = lambda x: x
|
2000-07-02 20:00:07 +08:00
|
|
|
out = []
|
|
|
|
charmap = [0]*256
|
|
|
|
try:
|
|
|
|
for op, av in charset:
|
|
|
|
if op is NEGATE:
|
|
|
|
out.append((op, av))
|
|
|
|
elif op is LITERAL:
|
|
|
|
charmap[fixup(av)] = 1
|
|
|
|
elif op is RANGE:
|
|
|
|
for i in range(fixup(av[0]), fixup(av[1])+1):
|
|
|
|
charmap[i] = 1
|
|
|
|
elif op is CATEGORY:
|
|
|
|
# FIXME: could append to charmap tail
|
|
|
|
return charset # cannot compress
|
|
|
|
except IndexError:
|
|
|
|
# unicode
|
|
|
|
return charset
|
|
|
|
# compress character map
|
|
|
|
i = p = n = 0
|
|
|
|
runs = []
|
|
|
|
for c in charmap:
|
|
|
|
if c:
|
|
|
|
if n == 0:
|
|
|
|
p = i
|
|
|
|
n = n + 1
|
|
|
|
elif n:
|
|
|
|
runs.append((p, n))
|
|
|
|
n = 0
|
|
|
|
i = i + 1
|
|
|
|
if n:
|
|
|
|
runs.append((p, n))
|
|
|
|
if len(runs) <= 2:
|
|
|
|
# use literal/range
|
|
|
|
for p, n in runs:
|
|
|
|
if n == 1:
|
|
|
|
out.append((LITERAL, p))
|
|
|
|
else:
|
|
|
|
out.append((RANGE, (p, p+n-1)))
|
|
|
|
if len(out) < len(charset):
|
|
|
|
return out
|
|
|
|
else:
|
|
|
|
# use bitmap
|
|
|
|
data = []
|
|
|
|
m = 1; v = 0
|
|
|
|
for c in charmap:
|
|
|
|
if c:
|
|
|
|
v = v + m
|
|
|
|
m = m << 1
|
|
|
|
if m > MAXCODE:
|
|
|
|
data.append(v)
|
|
|
|
m = 1; v = 0
|
|
|
|
out.append((CHARSET, data))
|
|
|
|
return out
|
|
|
|
return charset
|
|
|
|
|
2000-06-29 16:58:44 +08:00
|
|
|
def _compile(code, pattern, flags):
|
2000-06-30 07:33:12 +08:00
|
|
|
# internal: compile a (sub)pattern
|
2000-06-30 00:57:40 +08:00
|
|
|
emit = code.append
|
2000-03-31 22:58:54 +08:00
|
|
|
for op, av in pattern:
|
2000-06-30 18:41:31 +08:00
|
|
|
if op in (LITERAL, NOT_LITERAL):
|
2000-06-30 15:50:59 +08:00
|
|
|
if flags & SRE_FLAG_IGNORECASE:
|
|
|
|
emit(OPCODES[OP_IGNORE[op]])
|
|
|
|
else:
|
|
|
|
emit(OPCODES[op])
|
2000-06-30 21:55:15 +08:00
|
|
|
emit(av)
|
2000-06-30 15:50:59 +08:00
|
|
|
elif op is IN:
|
|
|
|
if flags & SRE_FLAG_IGNORECASE:
|
|
|
|
emit(OPCODES[OP_IGNORE[op]])
|
|
|
|
def fixup(literal, flags=flags):
|
2000-06-30 21:55:15 +08:00
|
|
|
return _sre.getlower(literal, flags)
|
2000-06-30 15:50:59 +08:00
|
|
|
else:
|
|
|
|
emit(OPCODES[op])
|
2000-07-01 02:39:20 +08:00
|
|
|
fixup = lambda x: x
|
2000-06-30 15:50:59 +08:00
|
|
|
skip = len(code); emit(0)
|
2000-07-02 20:00:07 +08:00
|
|
|
for op, av in _charset(av, fixup):
|
2000-06-30 15:50:59 +08:00
|
|
|
emit(OPCODES[op])
|
|
|
|
if op is NEGATE:
|
|
|
|
pass
|
|
|
|
elif op is LITERAL:
|
|
|
|
emit(fixup(av))
|
|
|
|
elif op is RANGE:
|
|
|
|
emit(fixup(av[0]))
|
|
|
|
emit(fixup(av[1]))
|
2000-07-02 20:00:07 +08:00
|
|
|
elif op is CHARSET:
|
|
|
|
code.extend(av)
|
2000-06-30 15:50:59 +08:00
|
|
|
elif op is CATEGORY:
|
|
|
|
if flags & SRE_FLAG_LOCALE:
|
|
|
|
emit(CHCODES[CH_LOCALE[av]])
|
|
|
|
elif flags & SRE_FLAG_UNICODE:
|
|
|
|
emit(CHCODES[CH_UNICODE[av]])
|
|
|
|
else:
|
|
|
|
emit(CHCODES[av])
|
|
|
|
else:
|
|
|
|
raise error, "internal: unsupported set operator"
|
|
|
|
emit(OPCODES[FAILURE])
|
|
|
|
code[skip] = len(code) - skip
|
2000-06-30 18:41:31 +08:00
|
|
|
elif op is ANY:
|
|
|
|
if flags & SRE_FLAG_DOTALL:
|
2000-06-30 15:50:59 +08:00
|
|
|
emit(OPCODES[op])
|
2000-06-30 18:41:31 +08:00
|
|
|
else:
|
|
|
|
emit(OPCODES[CATEGORY])
|
|
|
|
emit(CHCODES[CATEGORY_NOT_LINEBREAK])
|
2000-06-30 15:50:59 +08:00
|
|
|
elif op in (REPEAT, MIN_REPEAT, MAX_REPEAT):
|
|
|
|
if flags & SRE_FLAG_TEMPLATE:
|
|
|
|
emit(OPCODES[REPEAT])
|
|
|
|
skip = len(code); emit(0)
|
|
|
|
emit(av[0])
|
|
|
|
emit(av[1])
|
|
|
|
_compile(code, av[2], flags)
|
|
|
|
emit(OPCODES[SUCCESS])
|
|
|
|
code[skip] = len(code) - skip
|
|
|
|
else:
|
|
|
|
lo, hi = av[2].getwidth()
|
|
|
|
if lo == 0:
|
|
|
|
raise error, "nothing to repeat"
|
|
|
|
if 0 and lo == hi == 1 and op is MAX_REPEAT:
|
|
|
|
# FIXME: <fl> need a better way to figure out when
|
|
|
|
# it's safe to use this one (in the parser, probably)
|
|
|
|
emit(OPCODES[MAX_REPEAT_ONE])
|
|
|
|
skip = len(code); emit(0)
|
|
|
|
emit(av[0])
|
|
|
|
emit(av[1])
|
|
|
|
_compile(code, av[2], flags)
|
|
|
|
emit(OPCODES[SUCCESS])
|
|
|
|
code[skip] = len(code) - skip
|
|
|
|
else:
|
|
|
|
emit(OPCODES[op])
|
|
|
|
skip = len(code); emit(0)
|
|
|
|
emit(av[0])
|
|
|
|
emit(av[1])
|
2000-07-04 05:31:48 +08:00
|
|
|
mark = MAXCODE
|
|
|
|
if av[2][0][0] == SUBPATTERN:
|
|
|
|
# repeated subpattern
|
|
|
|
gid, foo = av[2][0][1]
|
|
|
|
if gid:
|
|
|
|
mark = (gid-1)*2
|
|
|
|
emit(mark)
|
2000-06-30 15:50:59 +08:00
|
|
|
_compile(code, av[2], flags)
|
|
|
|
emit(OPCODES[SUCCESS])
|
|
|
|
code[skip] = len(code) - skip
|
|
|
|
elif op is SUBPATTERN:
|
2000-07-04 05:31:48 +08:00
|
|
|
gid = av[0]
|
|
|
|
if gid:
|
2000-06-30 15:50:59 +08:00
|
|
|
emit(OPCODES[MARK])
|
2000-07-04 05:31:48 +08:00
|
|
|
emit((gid-1)*2)
|
2000-06-30 15:50:59 +08:00
|
|
|
_compile(code, av[1], flags)
|
2000-07-04 05:31:48 +08:00
|
|
|
if gid:
|
2000-06-30 15:50:59 +08:00
|
|
|
emit(OPCODES[MARK])
|
2000-07-04 05:31:48 +08:00
|
|
|
emit((gid-1)*2+1)
|
2000-06-30 18:41:31 +08:00
|
|
|
elif op in (SUCCESS, FAILURE):
|
|
|
|
emit(OPCODES[op])
|
2000-07-04 02:44:21 +08:00
|
|
|
elif op in (ASSERT, ASSERT_NOT):
|
|
|
|
emit(OPCODES[op])
|
|
|
|
skip = len(code); emit(0)
|
|
|
|
if av[0] >= 0:
|
|
|
|
emit(0) # look ahead
|
|
|
|
else:
|
|
|
|
lo, hi = av[1].getwidth()
|
|
|
|
if lo != hi:
|
|
|
|
raise error, "look-behind requires fixed-width pattern"
|
|
|
|
emit(lo) # look behind
|
|
|
|
_compile(code, av[1], flags)
|
|
|
|
emit(OPCODES[SUCCESS])
|
|
|
|
code[skip] = len(code) - skip
|
|
|
|
elif op is CALL:
|
2000-06-30 18:41:31 +08:00
|
|
|
emit(OPCODES[op])
|
|
|
|
skip = len(code); emit(0)
|
|
|
|
_compile(code, av, flags)
|
|
|
|
emit(OPCODES[SUCCESS])
|
|
|
|
code[skip] = len(code) - skip
|
|
|
|
elif op is AT:
|
|
|
|
emit(OPCODES[op])
|
|
|
|
if flags & SRE_FLAG_MULTILINE:
|
2000-07-01 06:37:31 +08:00
|
|
|
emit(ATCODES[AT_MULTILINE.get(av, av)])
|
2000-06-30 18:41:31 +08:00
|
|
|
else:
|
|
|
|
emit(ATCODES[av])
|
|
|
|
elif op is BRANCH:
|
|
|
|
tail = []
|
|
|
|
for av in av[1]:
|
2000-07-24 05:46:17 +08:00
|
|
|
emit(OPCODES[op])
|
2000-06-30 18:41:31 +08:00
|
|
|
skip = len(code); emit(0)
|
2000-07-24 05:46:17 +08:00
|
|
|
emit(MAXCODE) # save mark
|
2000-06-30 18:41:31 +08:00
|
|
|
_compile(code, av, flags)
|
|
|
|
emit(OPCODES[JUMP])
|
|
|
|
tail.append(len(code)); emit(0)
|
|
|
|
code[skip] = len(code) - skip
|
|
|
|
emit(0) # end of branch
|
|
|
|
for tail in tail:
|
|
|
|
code[tail] = len(code) - tail
|
|
|
|
elif op is CATEGORY:
|
|
|
|
emit(OPCODES[op])
|
|
|
|
if flags & SRE_FLAG_LOCALE:
|
|
|
|
emit(CHCODES[CH_LOCALE[av]])
|
|
|
|
elif flags & SRE_FLAG_UNICODE:
|
|
|
|
emit(CHCODES[CH_UNICODE[av]])
|
|
|
|
else:
|
|
|
|
emit(CHCODES[av])
|
2000-07-04 05:31:48 +08:00
|
|
|
elif op is GROUPREF:
|
2000-06-30 18:41:31 +08:00
|
|
|
if flags & SRE_FLAG_IGNORECASE:
|
|
|
|
emit(OPCODES[OP_IGNORE[op]])
|
|
|
|
else:
|
|
|
|
emit(OPCODES[op])
|
|
|
|
emit(av-1)
|
2000-07-03 01:33:27 +08:00
|
|
|
elif op in (MARK, INDEX):
|
2000-06-30 18:41:31 +08:00
|
|
|
emit(OPCODES[op])
|
|
|
|
emit(av)
|
2000-06-30 15:50:59 +08:00
|
|
|
else:
|
|
|
|
raise ValueError, ("unsupported operand type", op)
|
2000-03-31 22:58:54 +08:00
|
|
|
|
2000-06-30 07:33:12 +08:00
|
|
|
def _compile_info(code, pattern, flags):
|
|
|
|
# internal: compile an info block. in the current version,
|
2000-07-02 20:00:07 +08:00
|
|
|
# this contains min/max pattern width, and an optional literal
|
|
|
|
# prefix or a character map
|
2000-06-30 07:33:12 +08:00
|
|
|
lo, hi = pattern.getwidth()
|
|
|
|
if lo == 0:
|
2000-06-30 15:50:59 +08:00
|
|
|
return # not worth it
|
2000-06-30 07:33:12 +08:00
|
|
|
# look for a literal prefix
|
|
|
|
prefix = []
|
2000-07-02 20:00:07 +08:00
|
|
|
charset = [] # not used
|
2000-06-30 07:33:12 +08:00
|
|
|
if not (flags & SRE_FLAG_IGNORECASE):
|
2000-06-30 15:50:59 +08:00
|
|
|
for op, av in pattern.data:
|
|
|
|
if op is LITERAL:
|
2000-06-30 21:55:15 +08:00
|
|
|
prefix.append(av)
|
2000-06-30 15:50:59 +08:00
|
|
|
else:
|
|
|
|
break
|
2000-06-30 07:33:12 +08:00
|
|
|
# add an info block
|
|
|
|
emit = code.append
|
|
|
|
emit(OPCODES[INFO])
|
|
|
|
skip = len(code); emit(0)
|
|
|
|
# literal flag
|
|
|
|
mask = 0
|
2000-07-02 20:00:07 +08:00
|
|
|
if prefix:
|
|
|
|
mask = SRE_INFO_PREFIX
|
|
|
|
if len(prefix) == len(pattern.data):
|
|
|
|
mask = mask + SRE_INFO_LITERAL
|
|
|
|
elif charset:
|
|
|
|
mask = mask + SRE_INFO_CHARSET
|
2000-06-30 07:33:12 +08:00
|
|
|
emit(mask)
|
|
|
|
# pattern length
|
2000-07-02 20:00:07 +08:00
|
|
|
if lo < MAXCODE:
|
|
|
|
emit(lo)
|
|
|
|
else:
|
|
|
|
emit(MAXCODE)
|
|
|
|
prefix = prefix[:MAXCODE]
|
|
|
|
if hi < MAXCODE:
|
2000-06-30 15:50:59 +08:00
|
|
|
emit(hi)
|
2000-06-30 07:33:12 +08:00
|
|
|
else:
|
2000-06-30 15:50:59 +08:00
|
|
|
emit(0)
|
2000-06-30 07:33:12 +08:00
|
|
|
# add literal prefix
|
|
|
|
if prefix:
|
2000-07-02 20:00:07 +08:00
|
|
|
emit(len(prefix))
|
|
|
|
if prefix:
|
|
|
|
code.extend(prefix)
|
|
|
|
# generate overlap table
|
|
|
|
table = [-1] + ([0]*len(prefix))
|
|
|
|
for i in range(len(prefix)):
|
|
|
|
table[i+1] = table[i]+1
|
|
|
|
while table[i+1] > 0 and prefix[i] != prefix[table[i+1]-1]:
|
|
|
|
table[i+1] = table[table[i+1]-1]+1
|
|
|
|
code.extend(table[1:]) # don't store first entry
|
|
|
|
elif charset:
|
|
|
|
for char in charset:
|
|
|
|
emit(OPCODES[LITERAL])
|
|
|
|
emit(char)
|
|
|
|
emit(OPCODES[FAILURE])
|
2000-06-30 07:33:12 +08:00
|
|
|
code[skip] = len(code) - skip
|
|
|
|
|
2000-07-24 05:46:17 +08:00
|
|
|
STRING_TYPES = [type("")]
|
|
|
|
|
|
|
|
try:
|
|
|
|
STRING_TYPES.append(type(unicode("")))
|
|
|
|
except NameError:
|
|
|
|
pass
|
|
|
|
|
2000-06-02 01:39:12 +08:00
|
|
|
def compile(p, flags=0):
|
2000-06-30 00:57:40 +08:00
|
|
|
# internal: convert pattern list to internal format
|
2000-06-30 07:33:12 +08:00
|
|
|
|
|
|
|
# compile, as necessary
|
2000-07-24 05:46:17 +08:00
|
|
|
if type(p) in STRING_TYPES:
|
2000-06-30 15:50:59 +08:00
|
|
|
import sre_parse
|
|
|
|
pattern = p
|
2000-07-01 06:37:31 +08:00
|
|
|
p = sre_parse.parse(p, flags)
|
2000-03-31 22:58:54 +08:00
|
|
|
else:
|
2000-06-30 15:50:59 +08:00
|
|
|
pattern = None
|
2000-06-30 07:33:12 +08:00
|
|
|
|
2000-06-02 01:39:12 +08:00
|
|
|
flags = p.pattern.flags | flags
|
2000-06-30 00:57:40 +08:00
|
|
|
code = []
|
2000-06-30 07:33:12 +08:00
|
|
|
|
|
|
|
# compile info block
|
|
|
|
_compile_info(code, p, flags)
|
|
|
|
|
|
|
|
# compile the pattern
|
2000-06-02 01:39:12 +08:00
|
|
|
_compile(code, p.data, flags)
|
2000-06-30 07:33:12 +08:00
|
|
|
|
2000-06-02 01:39:12 +08:00
|
|
|
code.append(OPCODES[SUCCESS])
|
2000-06-30 07:33:12 +08:00
|
|
|
|
2000-07-24 05:46:17 +08:00
|
|
|
# print code
|
|
|
|
|
2000-06-30 07:33:12 +08:00
|
|
|
# FIXME: <fl> get rid of this limitation!
|
2000-06-30 00:57:40 +08:00
|
|
|
assert p.pattern.groups <= 100,\
|
2000-06-30 15:50:59 +08:00
|
|
|
"sorry, but this version only supports 100 named groups"
|
2000-06-30 07:33:12 +08:00
|
|
|
|
2000-07-03 06:25:39 +08:00
|
|
|
# map in either direction
|
|
|
|
groupindex = p.pattern.groupdict
|
|
|
|
indexgroup = [None] * p.pattern.groups
|
|
|
|
for k, i in groupindex.items():
|
|
|
|
indexgroup[i] = k
|
|
|
|
|
2000-06-02 01:39:12 +08:00
|
|
|
return _sre.compile(
|
2000-07-04 02:44:21 +08:00
|
|
|
pattern, flags, code,
|
|
|
|
p.pattern.groups-1,
|
|
|
|
groupindex, indexgroup
|
2000-06-30 15:50:59 +08:00
|
|
|
)
|